Although treatment for children with high-risk metastatic neuroblastoma has improved significantly in the past 20 years, only 45% of these patients become long-term, disease-free survivors, and bone marrow is a frequent site of resistant or recurrent disease. Improvement in survival necessitates effective new therapies for patients based upon biologically based risk stratification at diagnosis and highly sensitive monitoring of response to therapy that will identify patient subgroups with different outcomes who currently are clinically indistinguishable. High-risk neuroblastomas have few recurrent somatic mutations, which obviates molecular classification and therapies that rely on frequently altered oncogenic drivers. We have developed novel and clinically applicable TaqMan(R) Low Density Array (TLDA) gene expression assays for prognostication at diagnosis and for quantifying tumor load in bone marrow and blood for this group of patients. Overall goal: Validate the clinical utility of these new molecular diagnostics in collaboration with the Childrens Oncology Group (COG) and then establish them as CAP/CLIA-accredited tests. Specific Aims: Confirm performance characteristics and clinical utility 1) of a prediction signature (14-gene TLDA assay) for untreated primary MYCN non-amplified neuroblastomas and 2) of a detection signature (5-gene TLDA assay) for quantifying tumor load in bone marrow and blood. 3) Determine if these assays alone or combined improve prediction of outcome in the context of imaging and clinical variables. Research Design and Methods: The performance characteristics and clinical utility of the TLDA assays for risk stratification and monitoring of neuroblastoma response will be validated, and tests will be installed in CAP/CLIA-accredited laboratories including those at Children's Hospital Los Angeles [CHLA] and Nationwide Children's/COG Biopathology Center [BPC] so that they may be used in the future for clinical decision-making. Analytical performance characteristics of these tests will be confirmed in both CHLA and BPC laboratories, and clinical utility studies will be performed at CHLA. The 14-gene prediction signature predicted outcome in three independent clinical trials, and we are confident that additional studies using 300 archived and prospective samples from four COG studies will replicate existing data. Initial studies of the 5-gene detection signature (CHGA, DCX, DDC, PHOX2B, and TH), which can identify one tumor cell among 106 normal bone marrow or blood cells, showed that tumor load in bone marrow and PBSC during induction, after myeloablative consolidation, and after post-consolidation therapy correlates with outcome. These results will be validated with 2,250 archived and prospective bone marrow, blood, and PBSC specimens from four COG studies. Comprehensive analyses aimed at optimizing stratification and monitoring for clinical decision making will include these molecular diagnostic tests, 131I-meta-iodobenzyguanidine (MIBG) imaging, and clinical variables. Summary: We anticipate that our moelcular diagnostic assays will improve prediction of outcome and development of more effective therapy for patients with high-risk metastatic neuroblastoma.